Karahan Tepe > a new cultural centre in the Urfa area in Turkey

Karahan Tepe settlement was first discovered in 1997, but was surveyed in 2000 and again in 2011 within the course of the Sanlıurfa City Cultural Inventory. The settlements are located within the boundaries of Sanlıurfa (ancient name Edessa), a city in southeast Turkey. The settlement lies on a plateau known as the Tektek Mountains (Tektek Dagları) 63km east of Sanlıurfa. Like Göbekli Tepe, Hamzan Tepe and Sanlıurfa Yeni Mahalle PPN (PrePottery Neolithic period) settlements located around the Harran Plain in the Urfa Region, Karahan Tepe is also a PPN settlement located on a high plateau foot on the eastern side of Harran Plain.


Introduction
Karahan Tepe settlement was first discovered in 1997, but was surveyed in 2000 and again in 2011 within the course of the Sanlıurfa City Cultural Inventory.The settlements are located within the boundaries of Sanlıurfa (ancient name Edessa), a city in southeast Turkey.The settlement lies on a plateau known as the Tektek Mountains (Tektek Dagları) 63km east of Sanlıurfa.Like Göbekli Tepe, Hamzan Tepe and Sanlıurfa Yeni Mahalle PPN (Pre-Pottery Neolithic period) settlements located around the Harran Plain in the Urfa Region, Karahan Tepe is also a PPN settlement located on a high plateau foot on the eastern side of Harran Plain.
During the surveys on Karahan Tepe in 2000, basinlike pools carved in bedrock and a considerable number of tools of flint, obsidian, river pebble, and limestone were discovered.The finds show that the settlement was in use in the Pre-Pottery Neolithic period.Furthermore, there were many T-shaped pil-lars, which are also familiar from Nevali Çori, Sefer Tepe, Hamzan Tepe and Göbekli Tepe.As a result of a new survey in 2011, a new snake-shaped relief was found carved on a T-shaped pillar, of which half was formerly revealed by illegal excavations.
The settlement at Karahan Tepe is located approximately 705 metres above sea level, at 39°30' 22" East, 37°08' 15" North.The settlement was located in a high region of the Tektek Mountains.The region constitutes the southeast end of the Urfa plateau and is a hilly region in terms of geomorphology, rather than being mountainous (Map 1).The limestone region is arid and heavily eroded.The settlement is located between two rocky hills (Map 2; Fig. 1).It extends over an area of 60 000 square metres 1 .Flint and stone limestone layers in molar shapes appear 1km east of the settlement.On the other hand, the nearest basalt source is 15km north of the site.
It is understood from an examination of the limestone and architectural finds that Karahan Tepe was used as a major settlement during the Pre-Pottery Neolithic period (Çelik 2000b.6-8).The only settlement to compare in dimensions with this major site which lies on the eastern edge of Harran plain is Göbekli Tepe, which is three times bigger (Beile-Bohn et al. 1998;Schmidt 1995).Karahan Tepe settlement was probably in a good location for setting traps and hunting, due to its proximity to Harran Plain, and having impassable valleys nearby in the Early Holocene period (Figs. 2 and 3).At the settlement, flint stone finds can be seen extensively over the whole area, numbering approximately 50 per 1m 2 .Karahan Tepe seems to have been inhabited only during the Pre-Pottery Neolithic period.Finds are categorised as architectural elements, finds with depictions and small finds.The small finds comprise tools made of flint and obsidian, chisels and adzes made of river pebble, beads, stone pot fragments, grind stones, and pestles.There are 617 small finds and 274 architectural finds; 266 of the architectural finds are in-situ T-shaped pillars (Çelik 2000b.4).

Architectural elements
In Karahan Tepe, there is a considerable number of architectural finds, many of which have been preserved in good condition.In the east of the settlement, almost the whole area apart from what was used for agriculture and which is full of rocks, is covered with T-shaped pillars which are still in-situ, with only the top 50-60cm visible on the surface (Fig. 4).These pillars were erected in rows, and spaced 1.5-2 metres apart.Some pillars revealed by illegal excavations are between 1 and 2m high, between 20 and 25cm thick, and between 90 and 50cm wide (Çelik 2000b.4-6).Many have deep flutes on their sides (Hauptmann 1991/92.28).The upper parts of the Neolithic wall and wall corners can be seen clearly in patches.Groups of circular cavities on rock surfaces with a diameter of 30cm and a depth of 10-15cm, which are also familiar from Göbekli Tepe and Hamzan Tepe, can also be seen here as examples of the technique of pool carving (Çelik 2010.Fig. 6;Beile-Bohn et al. 1998.Abb. 20) (Fig. 5).
A T-shaped rock part ready to be removed from the bedrock on the west foot of the settlement and on a 30 degree slope was found in what we think was a quarry for producing T-shaped pillars (Fig. 6).This pillar is approximately 4.5m high, 1.5m wide and 80cm thick.A similar pillar which is ready to be removed was found in Göbekli Tepe.This pillar outline, with a height of 7m and an upper part of 3m, is twice as big as the pillar outline at Karahan Tepe (DAI 1997.552, Fig. 1;Schmidt 1998.4).This is a quite practical technique, with roughly chiselling on the side of the rock to produce a T shape, whilst carving the other side 40cm wide and 1m deep.The solid pillar separated from the bedrock with this technique can easily be removed from the quarry also with the help of a slope.The quarry is located immediately on the western edge of the settlement.
In the settlement, there are no pillars higher than the one in the quarry.However, such large pillars may be found at lower levels.Still, it should be noted that the pillars found at Göbekli Tepe are also not close to this height of 7 metres.This shows that the pillars in the quarry were scaled down to 1/3 proportion.There is a high possibility of finding pillars with a height of 3m during excavations.Although no terrazzo floors have been found, they are expected to be found at lower levels.However, a staircase-like remnant was found on the east side of a plain rock which is still part of the settlement.

Reliefs and sculptures
The finds with depictions on limestone from the settlement consist of two snake reliefs carved on the sides of pillars, one piece from a carved and orna- mented pillar, one piece from a male statue, one bird-shaped relief, and three pieces of a stele with reliefs on which human-shaped arms can be seen.
In the east of the ruins, the farmland was also disturbed by illegal excavations.There is a pair of pillars, side by side, 1.5m apart (Fig. 7).On one side of one of these pillars, there is a 70cm snake relief (Figs. 8, 9 and 10) with a round head and curled body, which probably continued to the broken part of the pillar (Çelik 2000b.Fig.1).The snake relief on the short side of the pillar is similar to those found on totem poles in Göbekli Tepe, and the snakes on grooving stones and on flat stone plaques from Jerf el Ahmar (Stordeur et al. 1996.Fig. 5;Schmidt & Schmidt 2010.Fig. 1b;Schmidt 2010.Fig. 18).The dimensions of this pillar, the lower part of which is broken, are 1.28 x 75 x 21cm.Opposite the example with the snake relief is a second pillar -found in 2011 -with a barely distinguishable snake relief on one side (Fig. 11).Only the head of the snake in a triangular form is visible, as only the top part of the pillar was revealed by illegal excavations.Similar snake reliefs are known from the upper parts of pillars and statues in Göbekli Tepe and Nevali Çori settlements (Schmidt 2004.Abb. 3, 6;Hauptmann 1999a.Fig. 10).These two pillars are the same size as the pillars found at the 'Löwenpfeilergabäude' structure in Göbekli Tepe (Beile-Bohn et al. 1998).This shows that the pillars at Karahan Tepe might well be contemporary at least with the architecture of the Göbekli Tepe II layer.Furthermore, they are already the same size as the in-situ pillars revealed at the surface level at Göbekli Tepe (Çelik 2000b.7).
Another group of surface finds comprises animal figures carved on a piece of a pillar (Fig. 12a, b).On the existing 40cm surface of 86cm long stone, the head and forelegs of a rabbit, the hind legs and tail of a gazelle, and the hind legs of an unidentified animal are visible (Çelik 2000b.Fig.2).Another piece, part of a wide, fluted pillar, has an animal leg in relief on both sides -the only example of this type to be found in the region (Fig. 14).Due to the high relief figure of a lion on a pillar at Göbekli Tepe, found in 2006, the mystery of the pieces found at Karahan Tepe has been resolved (Schmidt 2008.Abb. 2g).These stones are probably fragments of reliefs made on pillars with high relief technique.Two similar stele pieces were also found with animal reliefs (Fig. 15).
Moreover, an anthropomorphic statue with a phallus, which has the same characteristics as a Yeni Mahalle sculpture, was found (Fig. 16).This new find is an interesting example in having the same features as statues found at Göbekli Tepe and Yeni Mahalle (Çelik 2000a.4-6;2007.Fig. 4).
There is also a fragment of a jar carved from limestone and familiar from Göbekli Tepe.

Technology and typology of lithics
Plenty of flint and a few obsidian finds were found at Karahan Tepe; most of these relate to the blade industry.The number of Palaeolithic finds is 591, of which 541 are flint, while 50 are obsidian (Tab.1).
The highest proportion of the flint assemblage are blades.Some are completely preserved, their length varies between 5, 8 and 2.8cm; the width between 1.8 and 1.1cm; and the thickness between 0.5 and 0.2cm.On the other hand, the length of flakes varies between 4 and 2.1cm, while width ranges between 3.7 and 2.5cm, and thickness between 1.4 and 0.5cm.
Overall, there are 193 pieces of debitage products, consisting of blades and flakes.The flint tools are represented by 294 pieces (54.3%) in the whole group of flint finds (Tab.1).There are also 25 pieces of manufacturing waste, 168 pieces of blades, 23 pieces of flakes, 30 pieces of cores and core renewing tools, and one piece of burnt flint among other flint finds.
As raw material, well-qualified flint stones were used.Surveys have revealed some flint deposits one kilometre east of the site.Only two bipolar cores have been found.
Regarding the range of colours of the flint finds, 224 are grey (41.4%); 60 are light grey (11.1%); 88 are dark grey (16.26%); 42 (7.78%)are brown; 101 (18.66%) are light brown, 18 are black (3.32%) and 8 (1.48%) are cream/beige.At Karahan Tepe, there were five pieces of cores which are all bipolar (Fig. 17.1-2).They frequently have the form of a naviform core.The existence of some blades that are shaped out of a unipolar core, show that bipolar cores were used at the site.How-ever, in the recent survey, no unipolar cores have been found.The blade flaking of bipolar cores varies between 1 and 1.5cm.Three pieces of five cores which have been examined are broken; only two are well-prepared.
Among the blades and tools with blade handles the proportions of trapezoid cross-section blades with blade blanks is 91%, while the proportion of triangular cross-section blades is 9%.
At Karahan Tepe, the total of 30 (5.5%) cores and core rejuvenation fragments is very rare in the whole group of flint finds.Only five (16.6%) are cores.Apart from these, 9 (30%) pieces of core renewing and 16 (53.4%)strap blade fragments were found (Fig. 17.3).Most of the cores are dark brown and grey; cores of similar colour were found at both Göbekli Tepe (Beile-Bohn et al. 1998.54) and Nevali Çori (Schmidt 1988.162).
There are 50 obsidian finds; as a tool, only one end scraper was found among them (Fig. 17.4).The other obsidian finds are blade fragments, flakes, and waste.The proportion of obsidian in the find group is 8.46% (Fig. 17.5-7).
Examining the flint tools in terms of typology, it is clear that there are tool types from the Pre-Pottery Neolithic.Arrowheads constitute the majority.The most significant of these are Byblos, Nemrik, and Aswad types.There is a total of 194 pieces of arrowheads; of this number 137 pieces are 'Byblos' type (71%); 2 are 'Aswad' type (1%); 33 are 'Nemrik' type (17%) and the remaining 22 (11%) are unidentified.
Upon examination, it is clear that there are similarities between the arrowheads dated from the late phase A to the mid-phase B of the Pre-Pottery Neolithic at the site (Cauvin 1994.78-95).Similar arrowheads were found at Göbekli Tepe (Schmidt 2001a.52, Fig. 10/3, Fig. 11/5) and Mureybet (Cauvin 1994.79)'Byblos' type arrowheads At Karahan Tepe, 'Byblos' type arrowheads are represented by 137 items.These have been assessed according to their retouched types as falling into six groups.Blades were taken from both bipolar and unipolar cores, due to the fact that some of the directions of the arrowheads

'Aswad' type arrowheads
At Karahan Tepe, only 2 (1%) pieces of such arrowhead type were found; one is complete, while of the other one, only the lower part remains (Fig. 21.5-6).

Unidentified arrowheads
Unidentified arrowheads are represented by 22 (11%) items.Considering their blade width and thickness, they are the upper parts of either arrowheads or spearheads, apart from one spearhead and one fragment, which is considered to be part of a spear.Apart from one, nothing is known about their forms.

Perforators
There are 20 (6.8%) perforators among the finds at Karahan Tepe.Four are intact and sixteen items are broken (Fig. 21.7-13).Some are quite large.The handles of all perforators are made of blade or small bodkins.Among the other tools, the proportion of perforators found in the second layer of Göbekli Tepe is approximately 10% (Schmidt 2001.51, Fig. 9;Beile-Bohn et al. 1998.59), and they show similarities with perforators found at Karahan Tepe.Also, similar small perforators to those found at Karahan Tepe were discovered in Çayönü's settings phase.The existence of perforators at Karahan Tepe, we believe, correlates with the refined bones and stone objects found at the site.Stone and bone knickknacks made with skilled workmanship are proof of this.

Hammer
Only one hammer tool (0.3%) was discovered at Karahan Tepe; the surface is worn from use, and one side is broken.Similar tools have been found at all Neolithic settlements.

End scrapers
There was a large number of end scrapers at Karahan Tepe -a total of 40 (13.6%), 4 of which are rounded.Others were in various shapes.End scrapers and round end scrapers are of the same dimensions.Both end scrapers and round end scrapers were produced with the same technology (Fig. 23.1-10).The reason for so many finds of end scrapers could be the T-shaped pillars found in-situ at Karahan Tepe.It is believed that end scrapers were the most appropriate tools for chiselling the pillars.The scrapers at Göbekli Tepe's three openings from the second phase, also with T-shaped pillars, comprised 11.2% of all tools.(Schmidt 2001.51, Fig. 9); at Sanlıurfa-Yeni Mahalle it was 6%.

Bifacial tool fragment
One bifacial tool fragment (0.3%) was found at Karahan Tepe.The length of this brown flint tool is 5.2cm, while it is 4.9cm wide and 3.6cm thick; most of the top part and a piece of the bottom part are broken.Also, a small part of one face is unpeeled.This bifacial tool can be dated to the 'Acheulian' Phases (Taskıran 2002.53).

Obsidian tools
There were fifty obsidian tools (8.1%).Among these, there was one end scraper (0.3%); the others are bodkins, small bodkins, chip pieces and waste from manufacturing (Fig. 17.4-7).The sizes of the obsidian bodkins vary from 3.5 to 0.9cm in length, 3 to 0.9cm in width, and 0.9 to 0.1cm in thickness.Nine are black, and 41 are made of translucent black obsidian.
There are also 7 (1.29%)coloured stone beads and 1 stone object in this group.Three of the stone beads are in good condition, while the others are broken (Fig. 24.5-6, 9).The stone object (Fig. 24.4) is cylindrical; similar ones were found at Ain Mallaha (Cauvin 1972.Fig. 23.2) and Shaar-Ha-Golan settlements (Cauvin 1972.Fig. 5.1, 23.2).Cauvin regarded stone objects in this form as stylised human figures deriving from a Natufian period tradition in the Levant (Cauvin 1972.20-31).

Concluding remarks
Karahan Tepe is a single period settlement onlythat is to say, in the Pre-Pottery Neolithic period, like Göbekli Tepe.This feature protected the settlement from later ravages, so that the settlement has lasted in-situ until the present.
Karahan Tepe's current appearance, with pillars in the ground, is very similar to the Göbekli Tepe II layer pillars.In the light of all the finds, it can be stated that Karahan

∴ ∴
pillars in-situ on the surface at Karahan Tepe have features which cannot be seen at any Neolithic settlement in the Sanlıurfa region.Fortunately, the pair of opposed pillars were thought by the local villagers to be tombs, so the villagers did not remove them or use the land for farming.
In a new visit under a survey project carried out in 2011 at the settlement, a tally-like description and a snake relief with a triangular head were also discovered (Figs. 11 and 13).This snake relief at Karahan Tepe can be compared to those found in layer III of Göbekli Tepe, so Karahan Tepe may be contemporary with Layer II of Göbekli Tepe.However, it may be possible to date this settlement back to phase A of the Pre-Pottery Neolithic after future excavations at Karahan Tepe.

Map 1 .
The site at Karahan Tepe.
Fig. 24.1, 3)  These arrowhead are in general classified as Byblos Type, and categorised as early Pre-Pottery Neolithic(Cauvin 1994.78-79,Fig. 24.2).'Nemrik' type arrowheadsAt Karahan Tepe, only 33 (17%) pieces of such arrowhead types were found.While there are 23 examples of intact arrowheads among these, the other ten pieces are broken 21.1-4)There are nine on which the direction of blade flaking of the arrowhead is single-track, whereas there are 24 bi-directional arrowheads.The arrowheads in this group are classified according to their retouched types into three groups.Blades were taken from both bipolar and unipolar cores, due to the fact that some of the directions of the arrowheads are singletrack, while others are multi-track.All the arrowheads between type I and type III show similarities with those from Göbekli Tepe (Beile-Bohn et al.